Genome duplication in amphibians and fish

32 Pages·2011·0.61 MB·English

Journal of Zoology
JournalofZoology. Print ISSN0952-8369
REVIEW
Genome duplication in amphibians and fish: an extended
synthesis
B.K.Mable1,M.A.Alexandrou2&M.I.Taylor2
1InstituteofBiodiversity,AnimalHealth&ComparativeMedicine,CollegeofMedical,Veterinary&LifeSciences,UniversityofGlasgow,
Glasgow,UK
2EnvironmentCentreWales,MolecularEcologyandFisheriesGeneticsLaboratory,SchoolofBiologicalSciences,CollegeofNaturalSciences,
BangorUniversity,Bangor,UK
Keywords Abstract
polyploidy;amphibians;fish;climatechange;
Wholegenomeduplication(leadingtopolyploidy)iswidelyacceptedasanimportant
unreducedgameteformation;phylogeny;
cytogenetics;genomesize. evolutionaryforceinplants,butitislessrecognizedasadriverofanimaldiversifica-
tion.Nevertheless,itoccursacrossawiderangeofanimals;thisreviewinvestigates
why it is particularly common in fish and amphibians, while rare among other
Correspondence
BarbaraK.Mable,InstituteofBiodiversity, vertebrates.Wereviewthecurrentgeographic,ecologicalandphylogeneticdistribu-
AnimalHealth&ComparativeMedicine, tionsofsexuallyreproducingpolyploidtaxabeforefocusingmorespecificallyonwhat
CollegeofMedical,Veterinary&Life factorsdrivepolyploidformationandestablishment.Insummary,(1)polyploidyis
Sciences,UniversityofGlasgow,Glasgow phylogeneticallyrestrictedinbothamphibiansandfishes,althoughentirefish,butnot
G128QQ,UK. amphibian,lineagesarederivedfrompolyploidancestors.(2)Althoughmechanisms
Email:[email protected] such as polyspermy are feasible, polyploid formation appears to occur principally
through unreduced gamete formation, which can be experimentally induced by
Editor:StevenLeComber temperature or pressure shock in both groups. (3) External reproduction and
fertilization in primarily temperate freshwater environments potentially exposes
Received10September2010;revised23 zygotestotemperaturestress,whichcanpromoteincreasedproductionofunreduced
February2011;accepted25April2011 gametes. (4) Large numbers of gametes and group breeding in relatively confined
areas could increase the probability of compatible gamete combinations in both
doi:10.1111/j.1469-7998.2011.00829.x groups. (5) Both fish and amphibians have a propensity to form reproductively
successfulhybrids;althoughtherelativefrequencyofautopolyploidyversusallopoly-
ploidy is difficult to ascertain, multiple origins involving hybridization have been
confirmedforanumberofspeciesinbothgroups.(6)Problemswithestablishmentof
polyploidlineagesassociatedwithminoritycytotypeexclusioncouldbeovercomein
amphibiansviaassortativematingbyacousticrecognitionofthesameploidylevel,
but less attention has been given to chemical or acoustic mechanisms that might
operate in fish. (7) There is no strong evidence that polyploid fish or amphibians
currentlyexistinmoreextremeenvironmentsthantheirdiploidprogenitorsorhave
broaderecologicalranges.(8)Althoughpathogens couldplaya roleintherelative
fitness of polyploid species, particularly given duplication of genes involved in
immunity,thisremainsanunderstudiedfieldinbothfishandamphibians.(9)Asin
plants,manyduplicatecopiesofgenesareretainedforlongperiodsoftime,indicative
ofselectivemaintenanceoftheduplicatecopies,butwefindnophysiologicalorother
reasonsthatcouldexplainanadvantageforallelicorgeneticcomplexity.(10)Extant
polyploidspeciesdonotappeartobemoreorlesspronetoextinctionthanrelated
diploidsineithergroup.Weconcludethat,whilepolyploidfishandamphibiansshare
anumberofattributesfacilitatingpolyploidy,cleardriversofgenomeduplicationdo
not emerge from the comparison. The lack of a clear association of sexually
reproducingpolyploidswithrangeexpansion,harshenvironments,orriskofextinc-
tion could suggest that stronger correlations in plants may be driven by shifts in
matingsystemmorethanploidy.However,insufficientdatacurrentlyexisttoprovide
rigoroustestsofthesehypothesesandwemakeapleaforzoologiststoalsoconsider
polyploidyasapossibilityincontinuingtaxonomicsurveys.
JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon 151
Polyploidyinamphibiansandfish B.K.Mable,M.A.AlexandrouandM.I.Taylor
Introduction Ourpurposeistoevaluatethephylogeneticandgeographic
distributions, hypothesized origins, reproductive biology
Genomesequencinghasrevealedthatacrossbothplantand and ecology of sexually reproducing polyploid fish
animalkingdoms,thevastmajorityofgenesareorganized andamphibians,tobetterunderstandthepotentialdrivers
inmultiplesetsratherthansinglecopies(Wolfe&Shields, of polyploidy. Kejnovsky, Leitch & Leitch (2009) recently
1997; Blanc et al., 2000; Edgell, Malik & Doolittle, contrasted differences between mammals and plants
2000; Donoghue & Purnell, 2005; Wessler & Carrington, that might make the former less prone to polyploidy,
2005). This extensive gene duplication is hypothesized to but our objective is to emphasize what features shared by
have arisen through multiple rounds of whole genome fish and amphibians might promote polyploid formation
duplication (WGD), and is thought to be fundamental for and establishment. We focus solely on bisexually repro-
speciation, diversification of gene functions and shaping ducing polyploids in order to avoid confounding environ-
genomic architecture across major eukaryotic groups mental and geographic effects related to breeding system
(Lynch,2002; Ramsey & Schemske, 2002; Blanc, Hokamp variation rather than polyploidy, which has been a
& Wolfe, 2003; Leitch et al., 2004; Wessler & Carrington, problem for interpreting patterns in plants, ostracods
2005; Chen, 2007). It is now accepted that two rounds of and insects (reviewed in Mable, 2003). We first set the
WGDoccurredduringtheearlydiversificationofchordates historical context for polyploid discoveries in fish and
and vertebrates, with strong evidence supporting a subse- amphibians,withanoverviewofthetypesofdatathathave
quent teleost fish-specific genome duplication (FSGD; been used to identify them and summarize the current
Dehal&Boore,2005;Hoegg&Meyer,2005;Putnametal., phylogenetic distribution of extant polyploid fish and am-
2008;Larhammaretal.,2009;VandePeer,Maere&Meyer, phibians, before focusing on factors that favour polyploid
2009).MorerecentWGDeventshavealsooccurred,result- formationandestablishment.
ing in species that are currently functionally polyploid. As Wethussurveyedthedistributionofpolyploidyinanurans
some degree of diploidization and loss of duplicate gene (frogsandtoads)comparedwithfishesandexploitedcompre-
expressioninevitablyfollowsWGD(Ferris&Whitt,1977d; hensivedatabasessummarizingtheirtaxonomyanddistribu-
Adams,2007;Flagel&Wendel,2010),itisoftendifficultto tion [Amphibian Species of the World, (Frost, 2010) http://
distinguish ‘ancient’ (i.e. paleopolyploid) from ‘recent’ research.amnh.org/vz/herpetology/amphibia/], ecology and
events.Nevertheless,inthisreviewwedefine‘extant’poly- life history (Amphibiaweb, http://amphibiaweb.org/) and
ploidspeciesasthosehavingtwicethechromosomenumber conservationstatus(InternationalUnionforConservationof
ofcloserelatives,sometimesretainingatleastsomepairing NatureRedlistofEndangeredspecies,http://www.iucnredlist.
ofmultiplechromosomecopiesduringmeiosisandretaining org). For fishes, genome size and karyotype data were
evidence of duplicate gene expression distributed through- obtained from the animal genome size database (Gregory
out the genome, with most having some recognizably 2005a)(http://www.genomesize.com/),whiletaxonomic,eco-
distinctivefeaturesfromtheircloselyassociateddiploids.It logicalandbiogeographicinformationwasminedfromFish-
isthedistributionsandcharacteristicsoftheseextantpoly- base (Froese & Pauly, 2008) (http://www.fishbase.org).
ploidsthatarethefocusofthisreview. Because ploidy state is not an attribute that is available for
Therearenocomprehensivesurveysoftheprevalenceof most species in these databases, we first compiled a list of
polyploidyinanimals,butithasbeendocumentedacrossa known extant polyploid species from the primary literature
wide range of taxa, including insects, crustaceans, mol- (please seesupporting informationTables S1 and S2, where
luscs, fish, amphibians, reptiles and (to a lesser extent) therelevantreferencesarecited),beforeusingthedatabasesto
mammals (reviewedinBogart, 1980; Lewis, 1980; Otto& update species designations and phylogenetic distributions,
Whitton, 2000; Le Comber & Smith, 2004; Gregory & examinegeographicdistributionsandecologicalpreferences,
Mable, 2005). Polyploidy is most common in organisms andassessendangeredspeciesstatus.
thatdonotregulatetheirinternaltemperature(i.e.plants We conclude that the most striking feature shared by
and ectothermic animals), and are therefore directly ex- polyploid fish and amphibians is external reproduction in
posedtochangesintheirenvironments.However,because freshwater environments, predominantly in regions where
polyploidy is not widespread among ectothermic verte- temperature fluctuations during the breeding season are
bratesingeneral,itbegsthequestionofwhysomegroups common.Itwouldthusbetantalizingtospeculatethatthis
are polyploid and others not. Although it is possible that could support previous hypotheses that rates of polyploid
intrinsicmechanismsregulatinggenomeintegrityconstrain formationandestablishmentareassociatedwithperiodsof
polyploid estabilshment, it is also possible that ecological climaticchangeand/orcurrentlyunstableorextremeenvir-
factors (i.e. living in habitats or conditions that favour onments(e.g.Hagerup,1932)becausepolyploidsaregeneti-
polyploidy),incombinationwiththeinherentlystochastic callymoreresilientthantheirdiploidprogenitorsorableto
nature of establishment of polyploid lineages [i.e. forma- exploitmoreextremehabitats.Wedonotfindevidencethat
tion in the midst of diploid progenitors (Levin, 1975; bisexual polyploids have broader ecological ranges or dis-
Husband, 2000); producing balanced chromosome sets] tributionsorareatlowerriskofextinctionthantheirdiploid
areresponsible. relatives but insufficient data currently exist to provide
In this review we question why, among vertebrates, robust tests and to fully understand the potential impacts
polyploidy is most frequent among fish and amphibians. ofclimatechangeonratesofpolyploidspeciation.
152 JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon
B.K.Mable,M.A.AlexandrouandM.I.Taylor Polyploidyinamphibiansandfish
Historicalcontext larger than expected genomic signatures of ancient WGD
events, which opens opportunities for studying changes in
While polyploid animals are now well documented, the gene expression following polyploidization. A special issue
existence of the first polyploid vertebrate (the Ambystoma onpolyploidyinNewPhytologist(Ainouche&Jenczewski,
jeffersonianum complex of salamanders) was not accepted 2010) emphasizes the contributions of rapid advances in
by the scientific community until 1964 (Uzzell, 1964). The genome sequencing technologies to understanding such
first polyploid frogs (Odontophrynus americanus and Cera- genomic consequences of polyploidy. However, even in
tophrys ornata) were described in 1966 (Saez & Brum- plants,westilldonothaveacompleteunderstandingofthe
Zorrilla,1966),butdespiteprovidingclearfiguresshowing factors that promote the formation and establishment of
multiple sets of chromosomes and multivalent formation polyploidy in the wild, the role that ecology plays in
during meiosis, the authors rather remarkably concluded polyploid speciation, and whether polyploidy accelerates
that ‘it does not mean that we believe in the existence of diversification rates or is an evolutionary dead end (re-
polyploidy’. Bogart (1967) later confirmed that these were viewedinLevin,2002;Soltis,Buggs,Doyleetal.,2010).
both octoploid species that reproduced bisexually. Earlier
research on fish also suggested that polyploidy played a
Identification ofpolyploids
major role in the evolution of the Salmonidae (Sva¨rdson,
1945) and the genus Coregonus (Kupka, 1948) but these Extant polyploids have been identified using a variety of
were discounted by some researchers, and polyploidy in techniques, including chromosome counts, detection of
Salmonidaeasof1967was‘notconsideredtobeabiological multivalent formation during meiosis, banding patterns in
fact’ (reviewed by Bogart, 1967). Despite the initial excite- markers such as allozymes, cell size comparisons and
mentoftheseearlydiscoveries,polyploidyhasneverreally genomesizeestimations.However,noneofthesetechniques
emergedtotheforefrontofattentionbyanimalevolution- are incontrovertible and controversies over polyploid
arybiologists. statusoftenremainunresolved[e.g.Viscacharat(Gallardo
Inathoroughandinsightfulreviewbasedonexperimen- et al., 2004; Svartman, Stone & Stanyon, 2005)]. This is
tal induction of polyploidy in amphibians compared with reflected in the wide range of estimates of polyploid
plants and insects, Gerhard Fankhauser (an embryologist frequencies in plants (Jenkins & Rees, 1991; Hilu, 1993;
from Princeton University) predicted that polyploidy was Masterson, 1994; Soltis & Soltis, 1999; Otto & Whitton,
likely to be evolutionarily important in vertebrates (Fan- 2000). Most recently, based on detailed phylogenetic
khauser,1945).Hesuggestedthat‘thefollowingdatashould comparisons, it has been estimated that 15% of flowering
begathered:(1)occurrenceandfrequencyofpolyploidyin plantand31%offernspeciationeventshavebeenaccom-
different groups of animals; (2) origin of these exceptional panied by a ploidy increase(Wood et al., 2009). However,
individuals, in other words, the mechanisms that are re- despite increased rigour of analytical methods, this study
sponsible for their production; (3) the effectiveness of reliedoninferringploidylevelprimarilyfromchromosome
methods for the experimental induction of polyploidy; (4) data,whichcanbeproblematic(Otto&Whitton,2000).No
thegeneraleffectsofpolyploidyoncellsize,bodysizeand singlemethodworksforallgroups,withdifferencesamong
viability,andonthegeneralphysiologyandbiochemistryof plants, fish and amphibians in the predominant methods
theorganism;(5)theoccurrenceofqualitativeeffects,which used to infer ploidy status. Most documented cases of
areaddedtothemoreobviousquantitativeconsequences’. polyploidy in animals have used a pluralistic approach,
Half a century later, the answers to Fankhauser’s queries rather than relying on a single method. Below we survey
remainlargelyunanswered.Thisispartlyaresultoftheview themajortypesofmethodsthathistoricallyhavebeenused
thatpolyploidycouldnotbeimportantinanimalsbecause: toidentifypolyploids.
(1)theyhavetoomuchdevelopmentalcomplexitycompared
withplants;(2)sexdetermining mechanismsinvertebrates
Cytogenetics
andDrosophila are expected to bedisruptedby changesin
dosage (Muller, 1925; Orr, 1990; Mable, 2004); (3) regula- The oldest methods for identifying polyploids are through
tionofbodysizeinthedevelopingovumcouldbealteredby cytogenetic assessment of chromosome numbers, banding
cellsizeincreasesanddosage(Manevto,1945). and meiotic configuration patterns. Theory suggests that
Thepossibility thatpolyploidy hasplayedanimportant allopolyploids segregate disomically, as they present two
roleinanimalevolutionhasrecentlyreceivedmoreattention setsofhomeologouschromosomesthatareunlikelytopair
(Donoghue&Purnell,2005;Gregory&Mable,2005;Volff, at meiosis, while autopolyploids segregate polysomically
2005) but discoveries of unrecognized polyploids remain becausetheirchromosomespairatrandomandformmulti-
rare, possibly because few researchers look for them, but valentsduringmeiosis(Chenuil,Galtier&Berrebi,1999).In
also due to the difficulty of identifying cryptic polyploids practice, the reestablishment of disomic inheritance in an-
(see‘IdentificationofPolyploids’).Thus,thereislikelytobe cient polyploids(deWet, 1980) orpolysomicinheritancein
a considerable underestimate of the distribution and fre- allopolyploids arising from close relatives sharing partly
quency of polyploidy in animals. In contrast, polyploid homologous chromosomes (Stebbins, 1950), means that
plants are the focus of modern genomic research not only there is likely to be a continuum between strictly disomic
duetotheireconomicimportance,butalsoduetothemuch andstrictlypolysomicinheritance(seereviewbySoltisetal.,
JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon 153
Polyploidyinamphibiansandfish B.K.Mable,M.A.AlexandrouandM.I.Taylor
2010).Ingeneral,thedemonstrationoftetravalentsduring Marker-basedmethods
meiosisandtetrasomicinheritanceprovidegoodevidenceof
polyploidstatusbutlackoftetravalentformationdoesnot Allozymes have been widely used to identify suspected
necessarily mean that a species is diploid, because even polyploids(basedonnumberofcopiesordosageofbands),
youngpolyploids experience some degreeof diploidization toassessmodesoforiginofpolyploidtaxa(allopolyploidor
(e.g. Le Comber et al., 2010). Additional difficulties arise autopolyploid)basedonpatternsofduplicategeneexpres-
whenchromosomalrearrangements(throughRobertsonian sion and sharing of alleles with putative ancestors, to
fissionsandfusions)leadtochangesinchromosomenumber identify genomic composition of hybrids and to make
that are not related to genome duplication. Because in- inferencesaboutfatesofduplicategenes.Becauseallozymes
creases in genomic content can be achieved through other comparetheproteinproductsofexpressedgenes,theyalso
meansthanduplication(particularlyinnoncodingregions), have been used to assess the degree of duplicate gene
genomesizealsoisnotalwaysareliableestimateofWGD. expression across loci and tissue types (Ferris & Whitt,
Genomesequencingstudieshaveconfirmedthattransposa- 1977d;Allendorf,1978;Bailey,Poulter&Stockwell,1978).
bleelements(whichcanresultinlargedifferencesingenome Infact,SusumoOhno(Ohno,Wolf&Atkin,1968)basedhis
sizes without changes in chromosome numbers) may con- precocious predictions that vertebrates had experienced
found relationships between DNA content and chromo- multiple rounds of WGD on patterns of duplicate gene
somenumbers(e.g.Parisodetal.,2010). expression in allozymes. Despite the advent of more ad-
Nevertheless,atleastinanurans,karyotypingremainsthe vanced technologies, allozymes remain one of the clearest
most reliable method of detecting polyploidy. This is facili- methods for cheaply and rapidly characterizing polyploid
tated by the conservation of basal chromosome number genomes.
(range 9–13) and the presence of large chromosomes that Because expertise in cytogenetics was also at its peak
allowbandingpatternsandrDNAdistributionstobeusedto amongevolutionarybiologistswhenallozymeswerepopular,
indicatechangesinchromosomemorphology.Infact,Bogart combiningthetwoapproachesislikelytohavebeenrespon-
(1991)notedthatspeciesgroupswhereRobertsonianchanges sible for a peak in discovery of new polyploid amphibian
inchromosomes(i.e.fusionsorfissions)arecommondonot species in the 1970s (supporting information Fig. S1). Allo-
tendtoincludepolyploidsandpolyploidspeciesoftensharea zymes have been used to characterize the cryptic parental
high degree of apparent synteny (based on chromosome origins of previously identified polyploid amphibians [e.g.
banding patterns) as their diploid progenitors. All known Tomopternatandyicomplex(Channing&Bogart,1996);Hyla
polyploid anurans have an even replication of chromosome versicolor complex (Ralin, 1978; Romano et al., 1987)], dis-
number compared with closely related diploids (supporting coverthatdisomic,tetrasomicandintermediatepatternscan
informationTableS1),oftenwithhighlysimilarchromosome be found within tetraploid families depending on the locus
morphologies, based on banding patterns and rDNA loca- and tissue type examined [Hy. versicolor (Danzmann &
tions(Bogart,1980;Sto¨cketal.,2005). Bogart,1983)],andtoidentifywhichsexualspeciescontribute
In fishes, the situation is more complicated (supporting to hybrid unisexual lineages [Ambystoma laterale complex
information Table S2). Recent analyses of whole genome (Bogartetal.,1985;Bi&Bogart,2006)].
sequencesandancestralreconstructionsuggestthattheances- Infishes,allozymeshavebeenusedtoidentifypolyploid
tor of all teleost fishes had a haploid chromosome comple- lineages[e.g.Samonidae(Allendorf&Thorgaard,1984)],to
ment of 12–13 chromosomes that increased to 23–24 inferhybridcomposition[e.g.tetraploidloaches(Slechtova
chromosomes after the teleost-specific duplication (Jaillon et al., 2003)], to identify duplicated loci based on tissue-
etal.,2004;Kasaharaetal.,2007;Nakatanietal.,2007).The specificexpressionpatterns(Ohnoetal.,1968;Ohno,1970,
modal diploid chromosome number for acanthopterygian 1993;Ferris&Whitt,1975,1977c,b,1978),toinferprefer-
fishesis48(Mank&Avise,2006),withcountsrangingfrom entialexpressionofparentalallelesinexperimentalcrosses
22to250andcountsof‘diploid’speciesrangingbetween22 [e.g. crosses between carp and goldfish (Danzmann &
and78.Genomesizeiscorrelatedwithchromosomenumber Down,1982)]andpreferentialpairingofhomeologues[e.g.
when all species are investigated, but this relationship is rainbow trout (Allendorf & Danzmann, 1997)], and to
weaker (although remains significant) when polyploids are assess the proportion of duplicate genes that remain ex-
removedfromtheanalysis.Mank&Avise(2006),suggestthat pressedin‘old’polyploids[e.g.intheSalmonidaec.50%of
seven to 20 polyploidization events have occurred in extant duplicated allozyme loci remain detectable (Allendorf &
ray-finned fish lineages based on analysis of chromosome Thorgaard, 1984); a similar proportion to other tetraploid
numbers,althoughthisisalmostcertainlyanunderestimate, fishes,whichretainbetween25and70%ofduplicatedloci
asmanypolyploidfisheshaveyettobekaryotypedandwithin (Li,1980)].Aninterestingpointtonoteisthatthestudyof
polyploidlineagesthereisoftenevidenceofmultipleindepen- the fate of duplicate genes based on allozymes in Cyprini-
dentduplications(M.A.AlexandrouandM.I.Taylor,pers. formes (Ferris & Whitt, 1977a) provided evidence for the
obs.).Nevertheless,chromosomenumbershavebeenusedto separationoffunctionofduplicatecopiesbytissuetypeor
identify polyploid fish species and even entire families. For developmentalstage(nowknownassubfunctionalization),
example, the Catastomidae are all tetraploids, with a basal alongwithevolutionofnewfunctionsorlossoffunctions.
chromosome complement of 100, which is twice that of DNA-basedmarkerssuchasmicrosatellitescanbeused
diploidcyprinids(Ferris,1984). directly to provide evidence of duplicate genes rather than
154 JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon
B.K.Mable,M.A.AlexandrouandM.I.Taylor Polyploidyinamphibiansandfish
duplication of expressed products. However, fluorescent forexample,therearetwooctoploidspecieswiththesame
peak heights or stained-band intensities in microsatellites chromosomenumber (8n=104) butone(Ceratophrysaur-
are not always directly proportional to dosage of starting ita) has a genome size of 6.34pg, whereas the other (Ce.
productsduetounevenamplificationofallelesinPCR,so ornata) has 13.4pg. Although a diploid or tetraploid pro-
that copy numbers can be more difficult to infer than for genitor has not been identified, a closely related diploid
allozymes. Nevertheless, the large numbers of alleles at (Ceratophryscalcarata)hasagenomesizeof2.3pg;neither
microsatellite loci may make dosage less important, as octoploidhas8(cid:3) itsDNAcontent.Thiscouldsuggestthat
polyploidycanbeidentifiedbythenumberofpeaksrather thegenomesizeoftheprogenitordiploidwasmuchhigher
than the differences in strength. Their higher levels of or that cryptic ancient polyploidy occurred in one ‘octo-
variation can also be useful for inferring parental origin ploid’ lineage, but might also be due to differences in
and segregation of alleles to assess inheritance patterns; in genome size estimation by different authors. For Xenopus,
carp,around60%ofmicrosatellite loci still amplifydupli- genome size estimates for tetraploid species range between
cate copies despite the duplication event having occurred 3.0and4.1pg,withanaverageratiocomparedwithSilurana
around12mya(Davidetal.,2003).Microsatellitesalsohave tropicalis (the only extant diploid species) of 2.03 (range
been used in conjunction with flow cytometry to establish 1.8–2.3).However,thisisnotthelikelyprogenitor,asithas
ploidy levels and provide evidence of occasional sex in a basal chromosome number of 2n=20 compared with
supposed unisexuals (Ramsden, Beriault & Bogart, 2006; 4n=36 in Xenopus. Again, higher ploidy levels show a
Bi,Bogart&Fu,2007,2009;Bogartetal.,2007). nonlinearincreaseinDNAcontent,withratiosof3.5inthe
octoploidXenopusvestitusandonly4.6inthedodecaploid
XenopusruwenzoriensiscomparedwithSi.tropicalis.
Genomesize
Fish genome size varies considerably, with C-values
Variation in genome size has the potential to uncover ranging between 0.35 and 132.83pg. C-values for teleosts
polyploidy,butformanyorganisms,thereisnorelationship range between 0.35 and 4.9pg, with the average around
betweenDNAcontentandchromosomenumber(Gregory, 1.2pg.However,themajorityofspecieshaveC-valuesinthe
2005a,b), so it is important to confirm estimates with range 0.5–2.0pg. The genome sizes of polyploid teleosts
chromosome counts. There are also a variety of methods rangefrom1.36to3.75pg,withameanof2.5pg(Smith&
available to estimate genome sizes (e.g. flow cytometry, Gregory, 2009). Smith & Gregory (2009) suggest that
feulgen image analysis densitometry) and there can be genome sizes are usually greater than 2.5pg if polyploidy
difficulties in calibrating estimates from different labora- hasoccurred.Usinggenomesizetoinferpolyploidyismore
tories(Hardie,Gregory&Hebert,2002;Gregory,2005a,b; complicated in more basal groups of fishes, in which
Leitch, 2007; Smith & Gregory, 2009). Methods that pro- estimates may be confounded by transposable elements.
vide relative, rather than absolute, measures of DNA con- Forinstancethecartilaginousfishes(Chondrychthyes)have
tent can be less problematic for identifying suspected C-values in the range 1.51–17pg; although polyploidy has
polyploids, particularly if they are conducted in the same beensuspectedtohaveplayedaroleintheirearlyevolution
laboratory. (Kendall et al., 1994; Stingo & Rocco, 2001), repetitive
In amphibians, absolute genome size is not a good elements are also likely to have led to increases (Olmo
predictorofploidylevel,particularlyinsalamanders,which et al., 1982; Kellogg et al., 1995) or decreases (Leitch &
have a large range in genome sizes, even among diploids Bennett,1997)ingenomesize.Moreover,chondrichthyans
(Gregory,2005a,b)(AnimalGenomeSizeDatabase,http:// andsarcopteryianshaveonlyfourHoxclusters(Venkatesh,
www.genomesize.com).Inanurans,usinggenomesizealone 2007; Amemiya et al., 2008; Putnam et al., 2008), whereas
toinferploidyisalsoequivocal.Diploidgenomesizesvary actinopterygians have seven or eight Hox clusters (Crow
widely,withdiploidsinthegenusHyla(c.5.0pg)havinga et al., 2006) which suggests that the large genome size of
genome size similar to tetraploid Neobatrachus, despite chondrichthyansmaynotbetheresultofpolyploidization.
having the same basal chromosome number (2n=24). At Insummary,comparinggenomesizetochromosomecounts
least in salamanders, differences in genome size of species canbeinformativeforinferringploidystatusinfishes,but
withthesamechromosomenumberhavebeenfoundtobe the more extensive rearrangements and lineage-specific
due to differences in repetitive DNA, rather than WGD polyploidy makes the classification of extant polyploidy
(Baldari&Amaldi,1976;Bozzoni&Beccari,1978).Insuffi- evenlesscertainthanforanurans(supportinginformation
cientgenomicdataexisttoassesswhetherthisisalsotruein TableS2).
anurans or whether lineage-specific paleopolyploidy has
also occurred. Comparisons of genome size between pre-
Cellsize
viouslyidentifieddiploidandtetraploidspeciespairsmaybe
moreilluminatingwithregardstothepotentialtousesuch Nucleussizeispositivelycorrelatedwithchromatinamount
data to infer ploidy levels. Data for the five diploid–tetra- in polyploids (reviewed in Manevto, 1945). Cell size, how-
ploid pairs where data are available show an average ever,isnotalwaysdirectlyproportionaltonuclearvolume.
tetraploid: diploid genome size ratio of 1.93 (supporting In yeast, for example, whereas haploid cells tend to be
informationTableS1).However,thesamedoesnotappear smallerthandiploidcellsunderreducednutrientconditions,
tobetrueforhigherploidylevels.InthegenusCeratophrys, they can be the same volume when grown in rich medium
JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon 155
Polyploidyinamphibiansandfish B.K.Mable,M.A.AlexandrouandM.I.Taylor
(Weiss,Kukora & Adams, 1975). The samecan betrue of berts, 1997b)]. In some frog species, differences in colour
higher ploidy levels (Mable, 2001) – for animals with patternsormorphometricmeasurementscanalsobeusedto
nucleated red blood cells, the erythrocyte volume of tetra- distinguish some species that differ in ploidy(e.g. Mahony,
ploidsisconsistentlyhigherthanindiploidsbutisoftenless Donnellan & Roberts, 1986; Castellano et al., 1998; Sto¨ck
than the factor of two that might be expected with the et al., 2005), but such differences are likely to have arisen
doublingofDNAcontent.Infrogs,tetraploidHy.versicolor from adaptation postspeciation, rather than the genome
(Hylidae) can be reliably distinguished from their diploid duplicationeventitself.
progenitors, Hyla chrysoscelis (Matson, 1990) based on Therehasbeenlessfocusinfishonusingmorphological
differencesinerythrocytesize,butthevolumeoftetraploid or behavioural cues to identify polyploids, but there is the
cellsislessthanthatexpectedfromtheDNAcontent(ratio possibilitythatdifferencesinquantityofchemicalproducts
1.3–1.5). This means that triploids can be difficult to thatareproducedindirectproportiontocellsizeorgenome
distinguish because the ranges in volume overlap (Mable, copy could be useful. As far as we are aware, no compar-
1989).ForoctoploidOd.americanus,erythrocytesizevaries isonshavebeenmadebetweenolfactorysignalcomponents
between juveniles and adults, with those of juveniles com- in relation to ploidy in fish or amphibians that primarily
parabletothoseofadulterythrocytesofthediploidOdonto- useodourcues.
phrynuscordobae(Grenatetal.,2009).
Infishes,genomesizeanderythrocytesizearealsocorre-
Identifyingalloversusautopolyploids
lated in both teleosts and cartilaginous fishes (Hardie &
Hebert,2003,2004).Interestingly,coldwaterfishhavelarger Determiningtheoriginofpolyploidorganisms,andwhether
cell sizes than warm water species when controlling for they have arisen via autopolyploidy or allopolyploidy is
genomesize(Hardie&Hebert,2003).However,distinguish- crucial toourunderstandingofpolyploid evolution.How-
ingwhetherthisisprimarilyduetopolyploidyornonduplica- ever, distinguishing between these mechanisms is difficult,
tion based genome expansion is difficult. Nevertheless, cell due to the continuum between disomic and polysomic
sizemaybeusefulfordifferentiatingbetweencloselyrelated inheritancethatexistsinmostpolyploidspecies,regardless
diploidandpolyploidspeciesorforms(Felipetal.,2001). ofwhethertheyarosethroughhybridizationorfromasingle
progenitorspecies.AsSoltisetal.(2010)eloquentlydiscuss,
there is also a different perspective between systematists
Phenotypiccharacteristics
(whoareinterestedinwhetherpolyploidsarosefromdiffer-
Polyploidy may lead to an increase in the overall size of ent species) and geneticists (who are interested in segrega-
organisms. Such ‘gigantism’ is prevalent among plants and tion patterns during meiosis). Nevertheless, considerable
insectsbutisnotapparentinfishandamphibians.Polyploidy effort has been devoted to testing polyploid origins based
infishesandamphibiansappearstoresultinareductionin oninheritancepatternsandmorerigorousstatisticalmeth-
thenumberofcells,sothateventhoughcellsizeisincreased, odologies are in development (reviewed by Soltis, Soltis,
overallbodysizeremainsthesameasindiploids(reviewedin Schemskeetal.,2007;Parisod,Holderegger&Brochmann,
Bogart,1980).Mostbisexually reproducing amphibians are 2010). For example, by incorporating Bayesian statistics,
foundincloseassociationwithrelatedandmorphologically Olson(1997)presentedamethodthatallowssimultaneous
similar diploids, which have often been implicated in their assessmentofdisomicversustetrasomicinheritance,rather
formation (reviewed by Bogart, 1980). Polyploid anurans than performing goodness of fit tests separately for each
tend tohave similar bodysizeas the diploids(Fankhauser, model.Basedononlytwoallozymeloci,theydemonstrated
1945;Bachman&Bogart,1975)anddonotseemtoexperi- disomic inheritance (allopolyploidy) in Astilbe biternata
enceradicalchangesinphysiology;forexample,Hy.versico- (Saxifragacea) using a very small sample size. For DNA-
lor havesimilar metabolic rates as theirdiploid progenitors basedapproachessuchasmicrosatellites,Bayesianmethods
(Kamel,Marsden&Pough,1985).However,species-specific havebeenproposedthatuselargenumbersofmicrosatellite
matingcallsusedbyfemalestochoosemateshavebeenused lociandlargenumbersofindividuals,toevaluatemodelsof
to identify cryptic polyploids, based on polymorphism in inheritance without use of progeny arrays (e.g. Catalan
mating calls among populations that were otherwise indis- etal.,2006).
tinguishable, with ploidy status subsequently confirmed by Moreflexiblestatisticalapproachesthatconsiderarange
allozymes and/or chromosome counts (Wasserman, 1970; ofinheritancepatternshavealsobeenproposed.Forexam-
Bogart & Wasserman, 1972; Vigny, 1979; Barrio, 1980; ple,basedonthenumberofhomoeologouscopiespresentin
Haddad,Pombal&Batistic,1994;Roberts,1997b;Sto¨ck& each species for a set of neutral markers, Chenuil et al.
Grosse, 2003). In grey treefrogs (Hy. versicolor complex), (1999) developed a method that does not require the
femalesprefercallsoftheirownploidylevel(Gerhardt,1974, assumption thatallopolyploidy leadstomultisomic inheri-
1982, 2005a,b; Klump & Gerhardt, 1987), and some char- tance. Based on this, using five microsatellite loci in eight
acters of the mating call (e.g. pulse rate) change as a direct cyprinid species(withoneto threerepresentatives of each)
consequence of the increase in cell size arising from poly- of the genus Barbus (Cypriniformes), the hypothesis that
ploidy(Bogart& Wasserman,1972;Bogart,1980;Keller& Europeantetraploidbarbsoriginatedthroughautopolyploi-
Gerhardt,2001;Hollowayetal.,2006),althoughthishasnot dywasrejected.Thesetestshaveprovenparticularlyuseful
been found in all polyploid groups [e.g. Neobatrachus (Ro- in cases where hybridization between particular taxa is
156 JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon
B.K.Mable,M.A.AlexandrouandM.I.Taylor Polyploidyinamphibiansandfish
known and well documented. Investigations of patterns of Anurans
inheritance within the polyploid cyprinid Cyprinus carpio
Because polyploid frogs live in close proximity to their
using 59 microsatellite markers suggested a hybrid origin
diploidrelatives(Bogart,1980),weprovidealistofbisexual
(Davidetal.,2003).Stiftetal.(2008)describealikelihood-
polyploids and their closely related diploids (supporting
based method incorporating intermediate inheritance pat-
information Table S1). Wherever possible, credit has been
terns, as well as more complicated patterns due to double
given to the authors who first described a species as
reduction, which provides a more realistic assessment of
polyploid.Becausethetaxonomyofmanyanuransremains
segregationpatternsinpolyploids.
controversial(e.g.Roelants,Gower,Wilkinsonetal.,2007;
Wiens, 2007), we provide the current classifications pro-
Currenttaxonomyandphylogenetic
videdintheFrostAmphibianspeciesoftheworlddatabase
distributionofpolyploids [largely based on the revised taxonomy provided in Frost
et al., (2006)], as well as the species definitions at the time
One feature shared by fish, amphibians and plants is their
thatpolyploidswereoriginallyidentified.
complex and dynamic taxonomic history. Ichthyologists,
Polyploidy has arisen independently in multiple amphi-
herpetologists and botanists historically have tended to
bian families. The majority of species in the basal family
share a passion for systematics, and it has been common
Pipidae are tetraploid or higher (Fig. 1). The model Pipid
forspeciestobereclassifiedmultipletimesnotonlyamong
speciesXenopuslaeviswasoriginallythoughttobediploid
species and genera but also among families. This has been
but it is now recognized as an ancient tetraploid, with
particularly true since molecular characters have been
extensive,butincompletediploidizationacrossmuchofits
widely used to resolve phylogenies; whole genome studies
genome (Kobel & Du Pasquier, 1986); octoploids and
willresultinfurtherrevisions.Itcan,therefore,bedifficult
dodecaploids also occur in the family (see Evans et al.,
tosortthroughoriginalreportsofpolyploidyinthefaceof
2004, 2005, 2008). Polyploidy has also been suggested in
changing taxonomy. Here we review the distributions of
otherbasal groups(Leiopelmatidae Green,Kezer& Nuss-
extantknownpolyploidanuransandfish,consideringchan-
baum,1984)andtheentireSirenidaefamilymaybeancient
ging species designations and relationships among families
polyploids (Morescalchi & Olmo, 1974), but these reports
orhigherlevelsofclassification.Asaresultofthisreview,it
remained unconfirmed. In the more derived groups
is interesting to note the absence of polyploid vertebrates
(e.g. Hylidae, Ranidae, Microhylidae), polyploid species
fromcertain regions. Notably,thereare norecordedpoly-
aremorescattered.Bisexuallyreproducingpolyploidanur-
ploidfishesinAustraliaorAntarctica(possiblyanartefact,
ans have been confirmed across eight traditional families
astheseareasareunderstudied),andincontrasttoplants,
(Gregory&Mable,2005)buttaxonomicrevisionssuggested
very few extend their ranges into polar and arctic areas.
in the Frost database (Frost, 2010) mean that the 43
Therearepolyploidfrogsacrossalltemperateandtropical
polyploids are now distributed across 12 families, with 19
continents, including Australia, but amphibians in general
inthefamilyPipidae(plussomeunnamedspecies);eightin
donotoccurintheAntarcticorArcticsotheirabsencethere
Leptodacylidae(nowdividedintofourfamilies);foureach
islesssurprising.
200 mya
Salamanders
(Jurassic)
130 mya
(Early Cret)
120 mya
(Early Cret)
65 mya
(K-T boundary)
Figure1Family-levelamphibiantree,redrawn
fromRoelantsetal.(2007),indicatingfamilies
where polyploid species are known and esti-
Neobatrachia (165 mya Jurassic)
mated dates of divergence of clades that in-
110 mya clude polyploid lineages. Suspected but
(Early Cret) unconfirmed cases of polyploid species have
beenreportedintheLeiopelmatidaeandSca-
phiopidae.TheentirefamilySirenidaehasbeen
reportedasancientlypolyploidbutthishasyet
tobeconfirmed.
JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon 157
Polyploidyinamphibiansandfish B.K.Mable,M.A.AlexandrouandM.I.Taylor
~280 mya
(Permian)
Teleostei
193 mya
(Jurassic)
Ostariophysi
128 mya
(Early Cret)
Figure2Phylogenyoffishorders,redrawnfrom
Santini et al. (2009), indicating orders where
polyploid species are known and estimated
dates of divergence. A single family in the
100 mya
Percomorphscontainspolyploidspecies(Chan-
(Early Cret)
nidae)butdivergenceofthislineagefromrela-
tives that do not include polyploids have not
been clearly established. There is a notable
Acanthomorpha absence of polyploids within the Acanthomor-
phaWithintheOstariophysi,spontaneouspoly-
ploids have been reported within the
GymnotiformesandCharaciformes.
in Myobatrachidae (now Lymnodynastidae) and Bufoni- is important to note that this is not an indication of when
dae;threeinMicrohylidae;twoinHylidae;andoneeachin polyploidspeciesarose,justwhenthefamiliesinwhichthey
Dicroglossidae, Arthroleptidae and Ranidae. Several new arefounddivergedfromfamilieswherepolyploidyhasnot
polyploidanuranspecieshavebeenreported(Chiasmocleis beenidentified.Theclusteringofdivergencetimesindiffer-
leucosticta,Cophixaluspansus,Scaphiophrynegottlebei,Cer- ent lineages that include polyploids could suggest that
atophyrs joazeirensis, Pleurodema cordobae) since the sum- climaticconditionsintheearlyCretaceousandbeginningof
maryinOtto&Whitton(2000),alongwithanumberofnew the Paleogene favoured speciation by polyploidy. It is intri-
Xenopus species (Evans et al., 2004) and the surprising guing to note that the Pipidae (African polyploids in the
finding of sexually reproducing triploid toads in the Bufo genus Xenopus) and Limnodynastidae (Australian poly-
viridiscomplex(Sto¨cketal.,2002,2006).Speciesstatushas ploidsinthegenusNeobatrachus)bothdivergedintheearly
also been given to diploids in what were previously mixed Cretaceous and both of these families include multiple
complexesofdiploidsandpolyploids(e.g.Od.americanus, polyploid species, some of which appear to have speciated
Phyllomedusaburmeistericomplex,Bu.viridiscomplex)and as polyploids orhave noextant diploidprogenitors (Mah-
a cryptic octoploid (Pleurod. cordobae) has recently been onyetal.,1986;Mable&Roberts,1997;Evansetal.,2004).
discoveredinpopulationsofPleurodemakriegi(seesupport- AmongtheRanoids(earlyCretaceousdiversification),poly-
inginformationTableS1forreferences). ploidsarefoundinfamilieswithineachofthemajorclades
Polyploidspecies(andtheirdiploidprogenitors)tendto (Ranids, Dicroglossids, Pyxicephalids, Arthroleptids and
be underrepresented in molecular phylogenies (e.g. Faivo- Microhylids), which could suggest that conditions favour-
vichetal.,2005;Frostetal.,2006),possiblyasaresultofthe ableforpolyploidyexistedbeforetheirdivergence.Particu-
practical difficulties of dealing with duplicated gene se- larly intriguing is the 65mya divergence (corresponding to
quences,butalsothefactthatpolyploidtaxaarenotstrictly the Cretaceous-Tertiary boundary) of the clade including
appropriate for phylogenetic analyses because they do not Hylids, Ceratophrids, Cycloramphids, Leptodacytlids and
arise by cladogenesis. This, combined with lack of knowl- Bufonids, all of which contain multiple independent di-
edge on the nature of origins for most species (auto or ploid–polyploid species pairs. It would be tempting to
allopolyploid; single vs. multiple), makes it difficult to speculatethattheenvironmentalinstabilityduringtheK–T
evaluate hypotheses about dates of origins of particular boundarypromotedpolyploidspeciation,ashasbeensug-
species pairs. There are also discrepancies between recent gested for plants (Fawcett, Maere & Van de Peer, 2009).
phylogenetic hypotheses for amphibians (e.g. Frost et al., Dates ofdivergence predictedfor Hy.versicolor (lastpost-
2006;Roelantsetal.,2007).However,Roelantsetal.(2007) glacial period, c. 12000–35000ya Otto et al., 2007), and
present an analysis of diversification rates and predicted species in the Pipidae (maximum 65mya and most species
divergence dates within amphibians that allows some in- thoughttohavearisenbeforethePleistocene(Evansetal.,
sights.Basedonthisfamily-levelphylogeny,weplottedthe 2004,2005)],correspondtotimeswhenclimaticconditions
phylogeneticdistributionoffamiliesthatincludepolyploids were likely highly variable. However, similarly to plants
andhighlighttheapproximatedatesofdivergenceofthose (Soltis,Soltis&Tate,2004),manyofthepolyploidanurans
thatincludepolyploidsfromtheirclosestrelatives(Fig.2).It arethoughttohavehadmultipleorigins(Ptacek,Gerhardt
158 JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon
B.K.Mable,M.A.AlexandrouandM.I.Taylor Polyploidyinamphibiansandfish
& Sage, 1994; Mable & Roberts, 1997; Espinoza & Noor, even functional octaploids with 500 chromosomes (Birstein
2002; Gerhardt, 2005b; Sto¨ck etal., 2005;Holloway et al., etal.,1997).Theyareconsideredancientrelicts,withstrikingly
2006) and so polyploidy may be an ongoing process. similarspeciespreservedinthefossilrecordasfarbackas200
Even if complete phylogenies or genomic evidence for million years (Bemis & Kynard, 1997). A study of genome
duplicate genes were available, dating origins precisely duplicationeventsandfunctionalreductionofploidylevelsin
thuscouldremaindifficult.AsDoyle&Egan(2010)point sturgeon has revealed that gene silencing, chromosomal re-
out, the divergence of homoeologous copies (duplicate arrangements and transposition events are likely to be the
copies from each progenitor parent) in an allopolyploid dominant mechanisms that have shaped Acipenseriforme
tracks the divergence of diploid species, not the origin of genomes(Ludwigetal.,2001).Withinthisstudy,microsatel-
thepolyploidandautopolyploidoriginscouldbeevenmore lite analyses show that the maximum ploidy level for the
difficult to infer, so that skepticism about estimated dates Acipenseriformes is tetraploid and not octaploid, conflicting
iswarranted. withoriginalestimates(Birsteinetal.,1997).Thesedifferences
maybeduetotheextinctionoftheoriginaldiploidAcipenser-
iforme ancestor, as the oldest extant genus within the order
Fishes
(Polyodon)containsspecieswhosechromosomescanbeeasily
For fishes, diploid ancestors of extant polploids are often arrangedintoquartets(suggestingthatPolyodonmayactually
not identifiable but, unlike amphibians, there are entire betetraploid).Furthermore,Acipensersturio(presumedtobe
polyploid families. Based on karyotyping and genome size diploidwith116chromosomes)hasaC-valueapproximately
analyses, extreme cytogenetic variation has been found doublethesizeofitsclosestrelatives.Thesebasalspeciesare
in certain lineages of fish (Hinegard, 1968; Hinegard & thoughttobeancienttetraploidsfunctioninginadiploidized
Rosen,1972; Venkatesh, 2003) and bisexually reproducing state(Vasil’ev,2009).Thus,thereareanumberofunresolved
extantpolyploidsoccurinawiderangeofactinopteriygiian issueswithinthisancientorderoffishesthatrequirefurtherin-
families, including the Acipenseridae (Birstein, Hanner & depth analyses, including complete taxonomic coverage,
DeSalle, 1997; Ludwig et al., 2001), Cyprinidae (David furtherkaryotypesandgenomesizeestimates.
et al., 2003), Catostomidae (Ferris, 1984), Callichthyidae TheorderCypriniformescontainsthelargestdiversityof
(Oliveira et al., 1992) and Salmonidae (Johnson, Wright fish polyploids known to date, with over 250 recognized
& May, 1987). Using the Cyprinidae and Salmonidae to polyploid species spread across North America, Europe,
illustrate attributes of polyploidy in fishes, Le Comber & Africa and Asia. Most cypriniform polyploids are tetra-
Smith (2004) conclude that polyploidy may have been of ploids or hexaploids, with chromosome complements ran-
considerableimportanceintheevolutionoffishes.Polyploi- gingbetween100and150,whileonespecies(Ptychobarbus
dyhasalsoevidentlyplayedaroleintheevolutionofsome dipogon) has an amazing 446 chromosomes. The largest
lungfish species (Lepidoseriniformes), particularly within family of freshwater fishes, the Cyprinidae (Teleostei:
thegenusProtopterus,withC-valuessurpassing80pg. Cypriniformes), contains many species and genera
Althoughfishphylogeniesalsoremainuncertain,weused withgreatcytogeneticvariation.Twootherfamiliescontain
the well-calibrated phylogeny presented in Santini et al. a large number of tetraploid species, notably the Catosto-
(2009)toillustraterelationshipsamongthemajorfishorders midaeandtheCobitidae;however,duetothecomplexityof
that include polyploids and plot rough divergence times Cypriniformsystematics,thisislikelytochangeinthefuture
(Fig. 2). One large assemblage of fishes (primarily fresh- (Ferris,1984;Suzuki&Taki,1996;Saitoh,Chen&Mayden,
water) referred to as the Ostariophysi encompass multiple 2010).Amolecularphylogeneticstudybasedoncytochrome
examplesofextantpolyploidfamilies(Cyprinidae,Catosto- b and rRNA sequence data has revealed that a single
midae,Cobitidae,Callichthyidae),constitutingasignificant polyploidization event occurred in the lineage leading to
proportionoftheknownpolyploidactinopterygiianfamilies theBotiinae(Cobitidae),suggestingasingle originforthis
within one massive clade. In addition, spontaneous monophyletictetraploidassemblage(Slechtovaetal.,2006).
polyploidshavebeenreportedinGymnotiformesandChar- Many cyprinid genera are composed of stable polyploid
aciformes, which are also within the Ostariophysi (Fig. 2). series,includingBarbus,Labeobarbus,Luciobarbus,Pseudo-
Oftheremainingpolyploidfishes,allarebrackish,anadro- barbus, Spinibarbus, Diptychus, Carrasius, Capoeta, Tor,
mous, or strictly freshwater species. With few exceptions, Cyprinus, Schizothorax, Sinocyclocheilus and more. The
polyploidyismorecommonamongearlydivergingteleosts genusBarbusconsistsisofparticularinterestasitcontains
and relict bony fishes than later diverging teleost lineages atleast350speciesandiswellknownforitsmorphological
such as the Perciformes (Leggatt & Iwama, 2003). Thus, variation, wide distribution and the existence of diploid,
after the original three rounds of ancient genome duplica- tetraploidandhexaploidspecies(Kloseetal.,1969;Berrebi,
tions, subsequent polyploidization events seem to have 1995;Machordom&Doadrio,2001).Inastudyfocusingon
occurred within particular families, while the majority of theevolutionaryhistoryandmodesofspeciationwithinthe
fishgenomesremainfunctionallydiploid. genusBarbus,Machordom&Doadrio(2001)reconstructed
The order Acipenseriformes (sturgeons) is distributed phylogenetic relationships based on three mitochondrial
acrossNorthAmerica,EuropeandAsia,andincludesspecies genes in an effort to infer patterns and processes of poly-
withmultiplelevelsofploidy,withdiploidspeciescontaining ploidy.Althoughtheauthorsfocusonsystematicrelation-
120 chromosomes, tetraploids with 250 chromosomes and shipswithinBarbus,theyproposethatgenomeduplication
JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon 159
Polyploidyinamphibiansandfish B.K.Mable,M.A.AlexandrouandM.I.Taylor
within this genus may be considered as a homoplasic Acipenseriformes, salmonids primarily occupy temperate
character, since it must have occurred over at least three regions; however, some species from the genus Coregonus
independent periods and/or in three independent African can also be found within polar climates. The extreme
regions. However, the lack of nuclear markers within this migratory behaviour of salmonids might be related to the
study make it difficult to cross-validate relationships and originalancestralshiftinploidylevel.Recentevidencefrom
potentiallyinferahistoryofhybridizationthroughtopolo- molecular phylogenetic analyses supports the Esociformes
gical incongruence. Another phylogenetic study relying (Esocidae and Umbridae composed strictly of freshwater
solely on cytochrome b comes to similar conclusions con- species) as the sister group of the Salmonidae (Broughton
cerningthemultipleoriginsofAfricanbarbs(Tsigenopou- etal.,2010).
losetal.,2002);yet,onceagain,theauthorsfocusmoreon Finally,severalspeciesfromthegenusChanna(Channi-
the systematic relationships of the genus rather than hy- dae) are the sole polyploid representatives of the order
pothesesdealingwithpolyploidizationevents.WithinAsia, Perciformes (Banerjee, Misra, Banerjee et al., 1988); an
the Yunnan province of China is home to an amazing extremely diverse lineage estimated to contain in excess of
diversity of cyprinid polyploids (Yu et al., 1987), but the 10000 species. Despite karyotypes of 78 and 104 chromo-
biogeography of these fishes remains understudied. Given somes, genome size estimates for these species are lacking,
thediversityofcypriniformpolyploids,combiningcompre- makingitdifficulttoaccuratelyassessploidystatus(Rishi&
hensive molecular phylogenies with more karytoyping and Haobam, 1984). It is very likely that a bias in taxonomic
genomesize estimatesand othertechniquesshould helpto sampling and a lack of more recent cytogenetic investiga-
resolve the complex history of genome duplications within tions have partially led to the pattern of imbalance in
thegroup. polyploidy among different taxonomic groups of fishes;
WithintheSiluriformes(Catfishes),thespeciesrichgenus and, given the diversity of Perciformes, species with dupli-
Corydoras (Callichthyidae: Corydoradinae) from the neo- catedgenomesmayyetbediscovered.
tropicalregionisthemostwellstudiedanddiversegroupof
polyploids, with an impressive range of genome sizes and Whatfactorsfavourpolyploid
karyotypic variability (Hinegard & Rosen, 1972; Oliveira
formation?
et al., 1988, 1992, 1993a,b; Fenerich, Foresti & Oliveira,
2004; Shimabukuro-Dias, Oliveira & Foresti, 2004). The While polyploidy is not quite as rare as often suggested in
latterresearchsupportstheexistenceofmultiplepolyploid animals,itisobviousthatextantpolyploidyisrestrictedto
groups within the genus, with chromosomes ranging be- particular groups. In this section, we focus on the drivers
tween 2n=44–132 and C-values between 0.65 and 4.5pg. that might promote formation of polyploid lineages,
The primary mechanisms presumed to have shaped the emphasizing traits shared by fish and amphibians. We
complex genomic variability within these lineage include specifically question whether there are intrinsic features or
Robertsonian translocations, fissions, fusions, inversions ecological preferences shared by fish and amphibians (or
andpolyploidyfollowedbyDNAloss(Oliveiraetal.,1992, at least those that give rise to polyploid lineages) that
1993). Despite the recent publication of a comprehensive make them more prone to the initial formation of poly-
molecular phylogenetic framework (Alexandrou et al., ploids. The majority of polyploidization events in both
2011), many species remain without cytogenetic informa- plants and animals are thought to have been the result
tion, making it difficult to infer ploidy levels within the of unreduced gamete formation (Winge, 1917; Hagerup,
genus. Given karyotypic variability, it is also possible that 1932; Rabe & Haufler, 1992; Ramsey & Schemske, 1998;
severalspeciesofHypostomus,Plecostomus,Trichomycterus Husband & Schemske, 2000; Ramsey, 2007) but other
andWallagoarepolyploid,butthisremainstobeconfirmed mechanisms, such as polyspermy are also possible. Unlike
(Rab, 1981; Fenerich et al., 2004). Other polyploid catfish plants, somatic polyploidization has not been described in
species that have been revealed at an intraspecific level animals. Under conditions where the frequency of these
include Heteropneustes fossilis (Pandian & Koteeswaran, types of events is highest, it seems probable that opportu-
1999)and Clarias batrachus (Mustafa &Shams, 1982),yet nitiesforformationofpolyploidswouldbemaximized.We
these are isolated examples in comparison to the genus alsoconsiderotherfactorsthatmightincreaseopportunities
Corydoras. for the formation of successful polyploid individuals, such
TetraploidyisanancestralconditionoftheSalmonidae, as gamete production, reproductive environment and pro-
originallyshownvialinkageanalyses(Johnsonetal.,1987) pensityforhybridization.
andmicrosatellitedata(Sakamotoetal.,2000;Gharbietal.,
2006), but also confirmed by PCR amplification of non-
Frequencyofunreducedgametes
orthologoussequencesindifferenttaxa(Angers,Gharbi&
Estoup,2002).TheSalmoniformesareallpolyploid,having Althoughunreducedgametesoccurspontaneouslyinmost
undergoneagenomeduplicationaftertheirseparationfrom vertebrates, theyappear to produceviable progeny mainly
the Esociformes, between 45 and 100mya (Allendorf & in ectotherms. Artificial experimentation suggests that the
Thorgaard,1984;deBoeretal.,2007;Santinietal.,2009). easeofunreducedgameteformationisparticularlyhighin
Most salmonids have a haploid C-value of c. 3pg, while fishandamphibians(Fankhauser,1945).Theabsenceofthe
karyotypes range between 2n=56–104. Similar to the pachytenecheckpoint,whichisameioticsurveillancesystem
160 JournalofZoology284(2011)151–182 (cid:2)c 2011TheAuthors.JournalofZoology (cid:2)c 2011TheZoologicalSocietyofLondon